Unable to send voltage through source more than that is available at gate of the Mosf

[Right_Then]

Dear Fellows,

i am very new to electronics and am not an engineer by trade only hobbyist. So pardon my ignorance. i am trying to build a buck converter in the bare minimum, for some reason thou i am only able to receive the voltage that i supply to the gate out of the source and not the voltage that i supply to the drain of the mosfet. i.e if i supply 12v at the gate of the mosfet and 30v at the drain output is only 9v and not 20-25v as expected while testing the mosfet if it is working or not.

i am not doing anything that a buck does but simply trying to test the mosfet that it works okay or not so i have yet not connected the inductor but simply attached a wire in place of it to complete the connection and checking all parts. but it seems that gate is working like drain and drain is working like gate. when i try it on breadboard it works. Only on pcb it seems to have problem.

i am merely connecting the gate of mosfet to a constant voltage of (12v,24v,30v ) and drain also to similar arbitrary voltages to see what would be the outcome. but voltage coming out of the source never seems to go beyond what is supplied at gate.

i am sure i am missing some silly observation to get it right but am unable to figure it out.
so please see the Attachments provided. if you could help with that.

Thanks
Regards

 

[cowboybob]

Welcome, Right_Then!

Rather than make us draw a schematic from your pictures, could you supply us with one of your circuit?

Sure would help...

 

[dougy83]

i am not doing anything that a buck does but simply trying to test the mosfet that it works okay or not so i have yet not connected the inductor but simply attached a wire in place of it to complete the connection and checking all parts. but it seems that gate is working like drain and drain is working like gate. when i try it on breadboard it works. Only on pcb it seems to have problem.

So the SD20N60 is an N-channel mosfet, right? For the mosfet to turn on, the gate voltage has to be more than the source. e.g. if the gate voltage is 12V, the source might be able to reach 9V.

For this reason, when using a NMOSFET as a high-side switch (i.e. switching the positive voltage source), there will be a bootstrap (or boost) capacitor (and some associated circuitry) to raise the gate voltage to perhaps 5-10V above the rail (which would be 35V in your case with a 30V supply).

Regarding your soldering, it looks a little unreliable - try to apply solder all around the pad instead of just tacking a single side of the component lead (this will lead to the solder disconnecting from the lead with any modest movement).

 

[alec_t]

I see no oscillator. What is supplying a switching voltage to the FET gate?

 

[Right_Then]

Thanks cowboybob for welcoming me
i never drew schematic before therefore i posted photos. well here is my attempt at that. you will see it below i hope.

dougy83 thanks for your observations. that means i will have to re-etch a pcb and attempt what you said
your statement is similar to what i read about h-bridge with n-mos at high end. i was presuming it maybe. yes i am
not that good at soldering yet. but why does one need capacitor in that bootstrap why cant we just connect source and gate together with diodes in-between and from middle of it supply the gate. see mstd2 attachment please.

alec_t thanks for replying . i will implement pwm through microcontroller later. it was only the buck part of the pcb there are three input terminals in it
one for gate voltage which will carry gate pwm . one for Drain voltage . one for output voltage.

Thanks to all
Regards

 

[kubeek]

the pwm signal needs to be referenced to the source of the mosfet, so you will need some kind of mosfet driver to achieve that. That diode thing in the second picture will definitely not work like you wish. Also that resistor between gate and ground needs to be between gate and source, otherwise you will exceed the mosfet´s maximum Vgs of +/-20V.

 

[audioguru]

I see a Mosfet source follower, not a buck voltage converter.
If the gate is +12V then the source will be about +2V to +7V.

The voltage input is not an oscillator so it is probably just +12VDC. Then the Mosfet is not switching on and off and it will get extremely hot if there is a load.

I could not find a datasheet for a SD20N60 Mosfet but there are many E-Bay (Chinese) ads on Google. Maybe it is obsolete.

 

[dougy83]

You might want to draw up a schematic of your idea before you etch a new board, just to get a quick proofread from someone.

As mentioned by others, the diode configuration you outlined above won't work, and will suffer from the same issue you are trying to troubleshoot.

 

[audioguru]

A buck voltage converter reduces a supply voltage by switching supply voltage pulses into a series inductor and filter capacitor.
The duty-cycle of the pulses determines the output voltage.

Here are two, a higher voltage one and a lower voltage one. They use a P-channel Mosfet (not your obsolete N-channel one) and the gate pulse voltages must be as shown.

 

[Right_Then]

audioguru kind of you to put together those schemes. Yes i do know what buck converter is as far as definition goes.
now i am trying to make it. i was trying to avoid p-mos because as far as i know they are not available in high ( voltage + amperage ) ( i maybe wrong ) and are slower than n-mos as they get bigger ( again i can be wrong ). i would ultimately want to bring down 250volts from mains to about <50volts with 20-30amps to spare. i put together this PCB in my enthusiasm that i may be able to do it. but n-mos bit me. may i know why you reversed the p-mos in lower voltages configuration. SD20n60 is not my deliberate choice. this is what was available in my markets local electronics store with 600volts tolerance. i too could not find its Datasheet. i am assuming it is similar to STB20nm60 because it did not burn when supplied with 30 volts at gate. another one with me is FQA38N30 of which datasheet is available.

kubeek thanks for reply you wrote "the pwm signal needs to be referenced to the source of the mosfet" what does that mean that Pwm needs to go to source instead of gate or source needs to be attached to the pwm signal along with gate some or how.

dougy83 thanks again now i am going to practice with certain configurations on breadboard to trouble you fellows again. especially that bootstraping thing it is just not entering my head. i would try to take it in.

Thanks all
Regards

 

[kubeek]

I think you´d better start with something less dangerous, say creating 24V from 48V or similar, it seems to me that you don´t yet have enough experience to be playing with mains directly. The danger is not just in the high potential that could kill you, also there is a high risk of fire and exploding components if you don´t do stuff right.

What I meant was, that to turn on a mosfet, you need around 12 volts more on gate than there is on source. The way you have it now, you have ~320V at the source, and something like 5 or 10V at the gate. That makes a nice -310V between source and gate, and that makes the mosfet blow up in a spectacular fashion, as the max permissible voltage is +/-20V. You need to use a proper driver arrangement to achieve normal operation, for example through an optocoupler.

 

[ronsimpson]

Here is a "high side MOSFET driver". There are many types. I have used this one. It is built to input a signal at ground. (TTL level) The output is gate to source, where the source can be any voltage from ground to 600 volts. So if the IC has a 15 volt supply the gate-source voltage will be 0 or 15 volts with out regard to the source voltage. Look at the examples. You need to keep a signal on the input. Do not go to 100% on. Keep less than 95% on.

 

[audioguru]

audioguru i was trying to avoid p-mos because as far as i know they are not available in high ( voltage + amperage ) ( i maybe wrong ) and are slower than n-mos as they get bigger ( again i can be wrong ).

Yes a P-channel is slower with less voltage and less current than N-channel.

I would ultimately want to bring down 250volts from mains to about <50volts with 20-30amps to spare.

Gasp!
250V? down to 50V at 30A? That is 1500W! To power a train?
Then you must use a high voltage NPN transistor or N-channel Mosfet to supply the gate of the P-channel Mosfet with +240V to +250V pulses.

May I know why you reversed the p-mos in lower voltages configuration.

I made a mistake and I deleted it. Your original circuit will not work.

 

[Right_Then]

kubeek

I think you´d better start with something less dangerous

lol... hahaha
i do wear rubber gloves thou. but noted. i am doing the same, that is why initially i was supplying only 30 volts.

audioguru

To power a train?

lol...
no just electrolysis. Train maybe in future. hahahaha

ronsimpson Thanks just what i needed. Thanks so much. are there any comparators for high voltages too??

Thanks All
Regards

 

[ronsimpson]

What do you need a comparator for?

 

[Right_Then]

ronsimpson i was just wondering if high voltages can be compared directly and logical action taken on them thereof

Thanks
Regards

 

[alec_t]

How are you proposing to isolate the 50V from mains? At the moment it looks like one of your 50V connections could be at mains potential so possibly LETHAL

 

[ronsimpson]

Here is a simple buck down. The 50V is not connected to the negative side like you drew but to the positive wire. The gate drive is now simple. N channel MOSFET works well this way. Current mode is safe. I did not include the feedback. I can include a simple way to monitor the 50V and send it into the feed back pin on the UC3842. (TL431 + opto isolator)

Look at the data sheet for the UC3842/43/44.

 

[Right_Then]

alec_t i was not intending to isolate anything i was just trying to see if the concept works and whether i can generate the pulse to get the desired output.
i created a boost converter before this. it converted my 18v - 3amp adjustable Tabletop power supply voltage to 250 volts by the use of a mini transformer by using the secondary side of transformer on the micro-controller side and primary on the output side for the lack of proper inductor ( i experimented). do you mean that kind of isolation. but it runs nothing just shows the voltage on multiplier. as soon as i put load on it voltage drops that is why i was trying my hand at buck first to test something at low voltages to find what i was doing right or wrong.

ronsimpsonThanks so much for those schemes. and those fan73611 and uc3842 parts. i would have never know fan73611 could handle such voltages.
because when i check for such parts on buying sites. i am checking the ratings like that of mosfet. it is clearly mentioned how much it can take like 100 or 600 max.
but for these parts minimum and maximum supply voltage is mentioned. which i have not seen in hundreds for instance fan73611`s is 10 and 20 it is only after opening its datasheet that i could see it can handle 600+ in its description. That is when you told it so. otherwise i would have just wondered how people make such things!!!.



Thanks to all
Regards

 

[alec_t]

i was not intending to isolate anything

At some point in time you really MUST, for safety's sake.